Material guide: musk ambrette — three fields saying don't use it, and a replacer that tops the corpus

About a 4 minute read.

The short version

• Name: musk ambrette, 2,4-dinitro-3-methyl-6-tert-butylanisole, CAS 83-66-9, 18 suppliers
• C12H16N2O5, molecular weight 268.27, occurrence reads not found in nature
• Odour: musty, sweet, ambrette, seedy; substantivity 400 hours (measured at 20% in benzyl benzoate)
• Three fields say don't use it: category, regulatory, cosmetic uses
• Zero appearances in 954 formulas; musk ambrette replacer appears 56 times
• Oral rat LD50 339 mg/kg, the lowest 6.7% of 1,282 fragrance materials

Three fields saying one thing

Most records do not oppose their own use this consistently:

Field	Content
category	information only not used for fragrances or flavors
regulatory	FEMA GRAS (withdrawn), CoE
cosmetic_uses	not used anymore
notes	widely used fragrance implicated in contact photosensitivity.

Category says information only, regulatory says GRAS withdrawn, uses says not used anymore, and notes gives the reason.

I have just swept the whole database: 6,457 records carry that "not used for fragrances or flavors" category, and this one is fourth by supplier count among them.

What it is

The head_synonym reads 2,4-dinitro-3-methyl-6-tert-butylanisole.

Unpacked: one benzene ring, one methoxy group, one tert-butyl group, and two nitro groups. Those nitro groups are the point: this is a nitro musk, the workhorse class of synthetic musks through the first half of the twentieth century.

And it does not exist in nature (occurrence: not found in nature). The "ambrette" in the name refers to the smell of ambrette seed oil, which it imitates rather than comes from.

What removed it was not acute toxicity

The toxicity_oral field reads an oral rat LD50 of 339 mg/kg.

Across the 1,282 materials that carry this value and have five or more suppliers:

	LD50 (mg/kg)
Median	2,840
Quartiles	1,400 – 5,000
musk ambrette	339

It sits in the lowest 6.7%. Low, certainly, but the actual bottom is sodium selenite (7), vanadium pentoxide (10) and cyclohexylamine (11). That is the poison bracket.

What removed it is the sentence in the notes field: contact photosensitivity.

In 1981 Giovinazzo and colleagues measured its action spectra in Photochem Photobiol, in both guinea pigs and humans (PMID 7267727). An action spectrum tells you which wavelengths trigger the reaction. It is the most direct kind of evidence in photoallergy work: not "people who used it developed a rash", but "here is what this molecule does under this band of ultraviolet".

It is the same class of problem as lemon oil's furocoumarins: put it on skin, go into sunlight, and it appears afterwards. The difference is that lemon oil has a furocoumarin-reduced version you can buy, and musk ambrette has no such version.

Zero in the corpus, and its replacer at the top

Across 954 formulas:

	Count
musk ambrette	0
musk ambrette replacer	56

Fifty-six is the most of any replacer in the corpus, ahead of bergamot oil replacer at 28.

The replacer's record (#26837, 15 suppliers, no CAS number):

	musk ambrette	Replacer
Odour	musty, sweet, ambrette, seedy	dry, ambrette, musk, sweet
Substantivity	400 hours @20%	342 hours @100%
CAS	83-66-9	(empty)
Category	information only	fragrance agents

Two odour words overlap (sweet, ambrette) and two do not. A replacer is not a copy.

Its median dose is 2.50%, quartiles 2.00–5.00%. Most frequent company: coumarin (31), dipg (27), hydroxycitronellal (23), bergaptene-reduced bergamot (23), phenethyl alcohol (21), and musk ketone (21).

Three nitro musks, three states

Material	Suppliers	Formula	Category	Regulatory	Cosmetic uses
musk ketone	43	C14H18N2O5	fragrance agents	(empty)	fragrance; perfuming agents
musk ambrette	18	C12H16N2O5	information only	GRAS withdrawn	not used anymore
musk xylol	20	C12H15N3O6	information only	(empty)	fragrance; perfuming agents

Three structurally similar nitro musks in three different states in one database.

And the musk xylol row contradicts itself: its category says not for fragrance use while its cosmetic uses field calls it a perfuming agent. I have done a good deal of checking one record's fields against each other; this is another instance, in text fields.

Musk ketone is entirely normal and still has 43 suppliers. It is the only one of the three still in use, and its 21 appearances in the replacer's co-occurrence table mean those formulas swapped out the ambrette and kept the ketone.

(In 2023 Wang and colleagues assessed the global distribution and ecological risk of synthetic musks in Environ Pollut, PMID 37245793. Nitro musks' environmental persistence is a separate thread from the skin problem.)

What are those 18 suppliers selling

Why does a material marked "not used anymore" still have 18 suppliers?

I have no answer, only possibilities: analytical reference standards, jurisdictions where it remains permitted, or a supplier count field that records history rather than the present.

This is a question the database cannot answer. The supplier count carries no timestamp.

What this doesn't establish

• I checked no jurisdiction's current regulations. Everything here is field content, and FEMA GRAS (withdrawn) describes US flavour status rather than cosmetic regulation.
• The photoallergy section cites a 1981 paper's title and the kind of method it used. I did not read it in full and quote none of its numbers.
• The LD50 ranking comes from the 1,282 materials with that field and five or more suppliers, not from the whole database.
• I do not know what is in the replacer. It has no CAS and no formula; the database supplies only an odour and a substantivity.
• I did not establish which side of the musk xylol contradiction is right.
• The paragraph about the 18 suppliers is speculation. I did not look at their catalogues.